From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 3357 invoked by alias); 2 Oct 2002 11:32:57 -0000 Mailing-List: contact gsl-discuss-help@sources.redhat.com; run by ezmlm Precedence: bulk List-Subscribe: List-Archive: List-Post: List-Help: , Sender: gsl-discuss-owner@sources.redhat.com Received: (qmail 3254 invoked from network); 2 Oct 2002 11:32:47 -0000 Received: from unknown (HELO mail.phy.duke.edu) (152.3.182.2) by sources.redhat.com with SMTP; 2 Oct 2002 11:32:47 -0000 Received: from lilith.rgb.private.net (rgb.adsl.duke.edu [152.16.67.32]) by mail.phy.duke.edu (Postfix) with ESMTP id 4F26330194; Wed, 2 Oct 2002 07:32:46 -0400 (EDT) Date: Wed, 02 Oct 2002 04:32:00 -0000 From: "Robert G. Brown" X-X-Sender: rgb@lilith.rgb.private.net To: Gert Van den Eynde Cc: gsl-discuss@sources.redhat.com Subject: Re: Laguerre Polynomial overflow... In-Reply-To: <20021002090340.413f6a83.gvdeynde@sckcen.be> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-SW-Source: 2002-q4/txt/msg00002.txt.bz2 On Wed, 2 Oct 2002, Gert Van den Eynde wrote: > > Dear Robert, > I cannot reproduce this using gsl-1.2 (my own compilation) on a SuSE > 8.0 linux machine. Could you try to compile and run the attached program > and see if that fails too? Before starting to mess with the internal code I upgraded to the gsl 1.1 I had available in the RH 7.3 RPM set. It already worked at that point. Obviously little brownies were at work removing odd convergence/overflow bugs in between...;-) The moral, I suppose, is that I should make sure I'm using the latest version before filing a bug report. I'm a Bad Dog... I appreciate it, though, and apologize for wasting your time. I'll definitely compile and work through 1.2 before worrying about my NEXT gsl problem -- the lack of (visible) adaptive stepsizes in the gsl ode solvers. Thus far I've had to solve on a very fine grid to get decent accuracy (bound state eigensolutions integrated out from the origin spend a bit of time crossing the axis where they oscillate and eventually become stiff), but from the manual it looks like some adaptive stepsize routines have already made their way into the later revisions. I can't wait to try them out. BTW, the GSL is a fabulous toolset and I don't mean to sound critical at all. NR sucks (yes, I've read the rant, and then there is the strangulation noose -- I mean "copyright"). I'd like to contribute, if I ever have the chance. For example, I have (somewhere, I'd have to dig it out) a C translation of Forsythe, Malcom and Moler's "QUANC8" (8 panel Newton-Cotes quadrature) if that is of interest. I'm not really a computer scientist, but I've done a lot of numerical computations over the last decade or two. The other thing I might be able to contribute is at least simple genetic optimization algorithms and (if I could figure out how to frame them) neural network algorithms, as this is one of the things I do. rgb -- Robert G. Brown http://www.phy.duke.edu/~rgb/ Duke University Dept. of Physics, Box 90305 Durham, N.C. 27708-0305 Phone: 1-919-660-2567 Fax: 919-660-2525 email:rgb@phy.duke.edu